Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
on Carbon Emissions
Rahel Aichele
Gabriel Felbermayr
Abstract
Since 1997, CO2 emissions have continued to rise in many countries despite their emission caps under the Kyoto Protocol (Kyoto). Failure to meet promised targets, however,
does not imply that Kyoto has been pointless. Whether Kyoto has made a difference
relative to the counterfactual of No Kyoto is an empirical question that requires an
instrumental variables strategy. We argue that countries ratification of the statutes
governing the International Criminal Court is a valid instrument for ratification of Kyoto commitments. In our panel fixed effects estimations, the instrument easily passes
weak identification and overidentification tests. It can be plausibly excluded from our
second-stage equations and does not cause CO2 emissions. Our estimates suggest that
Kyoto ratification has a quantitatively large (about 10 percent) and robust, though
only moderately statistically significant, negative effect on CO2 emissions. We also
show that higher fuel prices and a different energy mix in Kyoto countries support this
C 2013 by the Association for Public Policy Analysis and Management.
result.
INTRODUCTION
Arguably, climate change is the most important global environmental problem of
our times. Its policy dimensions are explored by a large and insightful theoretical
literature. Due to the public goods nature of CO2 emissions, it is individually rational
for countries to free ride on others emission reductions. International environmental agreements (IEAs), such as the Kyoto Protocol, exist to solve this dilemma.
In the Kyoto Protocol (Kyoto), 37 industrialized nations and the European Union
(EU) have agreed to cap their levels of greenhouse gas (GHG) emissions to an
average of 94.8 percent of their 1990 emissions by the period 2008 to 2012. Yet since
1997, emissions have continued to rise in many countries despite their emission
caps. In 2010the latest year with GHG data available from the United Nations
Framework Convention on Climate Change (UNFCCC)many countries were still
far from achieving their promised GHG emission reductions. It appears as if the
Kyoto Protocol has been ineffective.
However, our main argument is that failure to meet a promised target does not
imply that Kyoto has been completely unsuccessful in bringing down emissions
relative to the counterfactual situation of No Kyoto (i.e., a counterfactual world
where no Kyoto Protocol exists). One needs to apply program evaluation techniques
to this large-scale policy intervention. Therefore, we ask whether there is empirical
Journal of Policy Analysis and Management, Vol. 32, No. 4, 731757 (2013)
C 2013 by the Association for Public Policy Analysis and Management
Published by Wiley Periodicals, Inc. View this article online at wileyonlinelibrary.com/journal/pam
DOI:10.1002/pam.21720
Taking into account the EU burden sharing agreement, the targets range from a 28 percent reduction
in Luxembourg to a 27 percent increase in Portugal.
2 The WTO is one rare example of an international agreement that provides some enforcement instruments (e.g., in the form of countervailing duties).
3 This question indeed arises for many international institutions and agreements, not just the ones on
global environmental goods.
RELATED LITERATURE
Our work is related to two important strands of literature. The first group of papers
deals with the economic rationale for IEAs and with their effectiveness. IEAs are
usually understood as ways to solve the free-riding problem inherent in climate
policy. The paper by Congleton (2001) offers a comprehensive survey. Andreoni and
McGuire (1993), Welsch (1995), Hoel (1997), Carraro and Siniscalco (1998), and
Lange and Vogt (2003) describe how international IEAs can solve this dilemma by
providing commitment devices and policy forums for coordination.
However, IEAs (like all international treaties) are based on the voluntary cooperation of countries. This raises the question of whether such agreements can be
Journal of Policy Analysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management
(1)
In our core exercise, Yit is the log of CO2 emissions. We also work with a broader
measure of all GHG emissions, but due to data availability, we present results as a
robustness check. Xit is a vector of controls that influence emissions. These controls
can be divided into three different categories. The first set of controls is comprised
of economic and demographic determinants of emissions (the log of GDP and its
square, the log of population, the log of the share of agriculture, manufacturing,
and services in GDP, and economic openness). We expect that an economically
large country has higher levels of emissions, all else equal; population growth increases emissions; and countries with a high share of manufacturing experience
higher levels of emissions, whereas a large share of services and agriculture in GDP
should be associated with fewer emissions. The second category includes measures
for preferences and policy (the stock of other IEAs, a countrys political orientation measured by the chief executives party affiliation, a dummy variable for WTO
membership, and the Polity index).5 The stock of other IEAs is a proxy for environmental awareness. So we expect a country that has signed up for more IEAs to
have lower emissions. Finally, Xit contains the spatial lag of Kyoto commitment,
that is, the Kyoto status of other countries weighted with their respective size over
distance squared. This measure reflects carbon leakage. We expect that countries
with large Kyoto countries nearby (i.e., with a larger value of the spatial Kyoto lag)
are less prone to competitiveness effects and thus have higher own emissions. To
account for time-invariant country-specific determinants such as endowments of
fossil fuels, patterns of comparative advantage, climatic and geographic conditions,
Aichele and Felbermayr (2012) differ from this study in several dimensions: first, the country coverage
is limited to only 40 countries; second, it is more narrowly focused on the proper calculation of carbon
footprints of countries and on their behavior relative to territorial emissions under Kyoto.
5 We do not include an EU dummy. In our fixed effects approach, only the change in EU status matters. In
our time period, with the exception of Malta and Cyprus, new EU members are EIT that have experienced
more changes than just a change in EU status. So, including an EU dummy would not be informative
about the effect of becoming an EU member country. However, in a robustness check we exclude
transition countries from the sample.
It implies that Kyoto starts to matter for committed countries once ratification
through the parliament has occurred. Studies evaluating the treatment effects of
IEAs such as the Montreal, Helsinki, and Oslo Protocol take a similar stance and
also use ratification as the decisive treatment date (see Ringquist & Kostadinova,
2005).
Econometric Issues
Our default strategy to eliminate the country-specific unobserved heterogeneity, i ,
from equation (1) is FE estimation on yearly data. It is fairly standard and used in
much of the empirical Kuznets curve and IEA treatment effects literature. Yet, this
strategy is subject to some criticism. Bertrand, Duflo, and Mullainathan (2004) argue that standard errors of treatment effects in FE estimation are inconsistent, and
the estimators standard deviation is underestimated if the outcome and treatment
variable are both serially correlated over time. This might lead to an over-rejection
of the Null of no effect. The authors suggest applying a FE estimator to the pretreatment and posttreatment averages (long FEs estimator) to cure the spurious
correlation problem. Although there has been some heterogeneity in the timing of
Kyotos ratification across countries, most countries ratified it between 2001 and
2003. So we assume treatment takes place in this period, but conduct robustness
checks pertaining to this choice. The pretreatment and posttreatment windows of
1997 to 2000 and 2004 to 2007 are chosen such that they constitute symmetric
intervals around this treatment window.6 Whereas the within transformation on
yearly data are subject to the Bertrand, Duflo, and Mullainathans (2004) critique, it
has the advantage that it does not require assumptions about a treatment window.
Moreover, the number of useable observations is about 10 times larger than in the
long FE model. For these reasons we show results for both methods.
FE estimation eliminates any correlation between the second-stage equation errors and the Kyoto dummy that would be due to time-invariant unobserved heterogeneity at the country level. However, such a correlation may arise for two further
reasons. First, if our Kyoto dummy is a noisy measure for a countrys true climate
policy stance, estimates based on a simple within-transformed model would be biased towards zero (attenuation bias). Second, a bias also arises in the presence of
reverse causation, that is, if a transitory shock on the outcome variable makes it
more or less likely that a country has a binding Kyoto commitment. Instrumenting
the Kyoto status cures these biases. Therefore, the key empirical challenge consists
in finding a valid instrument for Kyoto commitment.
Note that Russia and Ukraine ratified Kyoto in 2004 and Belarus in 2005 and are assigned to the
treatment group. Australia and Croatia ratified the Kyoto Protocol in 2007 and are assigned to the
control group.
(2)
Similarly, Mike Huckabee, former Governor of Arkansas, argues that the Kyoto Protocol would have
given foreign nations the power to impose standards on us (Huckabee, 2007, p. 70). China expressed
similar concerns in the Copenhagen climate change negotiations. Other developed or newly industrialized
countries that are neither a Kyoto nor an ICC member include Israel, Korea, Singapore, Chile, and Turkey.
8 This appears to be the natural choice since we are working with a linear fixed effects approach in the
second stage. However, the correlation also holds in a Probit framework.
Other international treaties, such as those governing the WTO or international environmental questions, cannot be easily excluded since they will affect emissions directly through green preferences of
voters and consumers, or through trade policy.
10 GHG data are available for the years 1990, 2000, 2005, and 2008. The country coverage ranges from
109 nations in 1990 to 134 for the latest data point.
11 The nonparametric time trend takes care of the global fuel price; country-level fixed effects take care
of endowment-related or geographic determinants of fuel price differences.
12 Pump prices of diesel and gasoline are only available every other year from 1998 to 2006.
13 The exact calculation of the spatial weighting matrix (e.g., with distance squared instead of simple
distance or with all others trade shares, i.e., exports plus imports, in GDP as a more direct measure
of economic competition) does not make a significant difference. Distance data come from the CEPII
distance database.
14 http://iea.uoregon.edu/.
Obs.
Mean
Std. dev.
Min
Max
Source
1,456
335
622
622
1,209
1,209
9.78
11.20
0.58
0.73
22.64
2.52
2.28
1.56
0.33
0.35
27.45
5.11
4.19
8.39
0.01
0.02
0.00
0.00
15.69
16.04
1.73
1.90
93.77
29.26
(a)
(a)
(a)
(a)
(a)
(a)
1,209
1,209
1,198
1,209
67.82
18.04
7.29
7.17
26.81
27.11
1.49
0.99
5.32 102.43
0.00 99.46
3.02 10.15
4.83
9.38
(a)
(a)
(a)
(a)
1,456
1,456
1,456
0.12
0.28
0.20
0.33
0.97
0.57
0.00
0.00
0.00
1.00
13.29
2.00
(b)
(b)
(b)
1,456
1,456
0.33
0.12
0.47
0.38
0.00
0.00
1.00
3.46
(c)
(c)
1,456 17.95
1,456 325.94
1,456
9.28
1,421
2.62
1,440
2.30
1,437
3.93
1,456
3.28
1,456
0.17
1,456
0.86
1,456
0.82
1,456
3.66
1.91
69.64
1.51
0.56
1.11
0.35
0.59
0.09
0.48
0.39
6.41
13.23 23.28
175.05 542.06
6.04 14.09
0.35
3.79
2.63
4.36
1.01
4.40
1.79
4.84
0.10
0.30
0.05
4.57
0.00
1.00
10.00 10.00
(d)
(d)
(d)
(a)
(a)
(a)
(e)
(f)
(d)
(g)
(h)
Note: The table shows summary statistics for variables over the period 1997 to 2007.
Sources: (a) World Bank WDI 2010, (b) http://www.unfccc.int, (c) UN Treaty Series database, (d) PWT
6.3, (e) http://iea.uoregon.edu, (f) World Bank DPI 2010 (series: execrlc), (g) http://www.wto.org, (h)
http://www.systemicpeace.org.
political authority characteristics and ranges from 10 to 10, where 10 is a perfectly autocratic regime and 10 is a full democracy. Table 1 lists summary statistics
and sources.15
The WDI database has information on emissions for almost 200 geographical
entities. However, we drop entries that do not constitute independent countries
(such as, e.g., Puerto Rico, Greenland, Monaco, and Hong Kong). Moreover, for a
number of countries, we do not have the full list of covariates as defined above. At
this stage we lose mainly small (island) states like Grenada, Barbados, but also some
very recently created countries (such as Kosovo). In our analysis, we drop countries
15
Additional information on Kyoto status, average emission growth from the pretreatment to posttreatment period, and sample info (rich, large, Organization of the Petroleum Exporting Countries (OPEC),
transition country) is given in Appendix Table A1. All appendices are available at the end of this article as
it appears in JPAM online. Go to the publishers Web site and use the search engine to locate the article
at http://www3.interscience.wiley.com/cgi-bin/jhome/34787.
16
Five committed Kyoto countries (Iceland, Liechtenstein, Luxembourg, Russia, and Switzerland) are
not included due to data availability.
We have also experimented with other international agreements such as the Comprehensive NuclearTest-Ban Treaty and the Anti-Personnel Land Mines Convention. Ratification of these texts also tends to
make Kyoto commitments more likely; however, the effects are weaker and less statistically significant.
18 Note that a positive and statistically significant effect is obtained when the spatial lag of ICC is not
included in the equation.
17
Model
Method
(1)
Excluded instruments
ICC (0,1)
ICC, spatial lag
(2)
0.19*** 0.11***
(0.05)
(0.04)
0.51***
(0.10)
Other controls
Kyoto, spatial lag
ln GDP
ln GDP, squared
ln population
ln manufacturing (percent of GDP)
ln agriculture (percent of GDP)
ln services (percent of GDP)
ln stock of other IEA
Government orientation (0.1, 0.2, 0.3)
Openness, (Exp + Imp)/GDP
WTO (0,1)
Polity (1 to 1)
No. of observations
Adj. R2
F-stat
Log-likelihood
2
1,456
0.25
4.32
1,456
0.52
6.69
Probit long
diff.a
(5)
(3)
Long FE
(4)
0.10***
(0.03)
0.41***
(0.08)
0.25***
(0.07)
0.37***
(0.12)
0.16***
(0.03)
0.04***
(0.01)
0.01
(0.01)
0.94*
(0.52)
0.02
(0.02)
1.72***
(0.38)
0.01
(0.04)
0.14**
(0.06)
0.01
(0.08)
0.02
(0.09)
0.03
(0.19)
0.22***
(0.07)
0.15**
(0.06)
0.00
(0.00)
1,418
0.61
7.36
0.01
(0.01)
2.46**
(1.02)
0.07**
(0.03)
1.31***
(0.45)
0.02
(0.10)
0.29**
(0.13)
0.15
(0.17)
0.27
(0.17)
0.07
(0.59)
0.27
(0.18)
0.28*
(0.14)
0.02
(0.01)
266
0.69
5.69
2.04***
(0.69)
0.05***
(0.02)
3.42***
(0.59)
0.16
(0.12)
0.05
(0.08)
0.25**
(0.13)
0.07
(0.11)
0.11
(0.20)
0.63***
(0.14)
0.16**
(0.07)
0.02***
(0.01)
133
13.92
39.60
Note: Linear probability and Probit models. Sample: 133 countries. Probit regression shows marginal
effects. Heteroskedasticity robust standard errors (clustered at country level) in parentheses. *P < 0.1;
**P < 0.05; ***P < 0.01. LPM: Year dummies and constant included (not shown).
a Estimation on cross-section of differences between pretreatment and posttreatment averages.
for Kyoto commitment: The higher the share of agriculture in a countrys GDP, the
lower the probability it will ratify the Kyoto Protocol. The logs of the manufacturing
and services shares are not relevant for Kyoto commitments.
We include two variables to proxy for green preferences. The first variable, the
log stock of other (than Kyoto) IEAs ratified by a country is expected to affect
the likelihood of commitment positively, but does not show up significantly in
the analysis. The second variable is the countrys chief executive partys political
Journal of Policy Analysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management
19
These countries are Albania, Armenia, Cambodia, China, Croatia, Estonia, Georgia, Jordan, Lithuania,
Latvia, Moldova, Macedonia, Nepal, Saudi Arabia, and Vietnam.
20 Time-differencing pre- and postpolicy averages, the selection equation yields a model in which the
dependent variable (Kyoto commitment) and the instrument (ICC) are again binary variables. This
procedure eliminates unobserved heterogeneity; yet, the Probit model remains applicable.
21 Note that Belarus is considered as treated, while Australia and Croatia are put in the control group.
.01
Density
.02
.03
.04
.05
50
25
25
50
Emission growth (in %)
Kyoto
75
100
NonKyoto
Note: The figure shows a kernel density plot (Epanechnikov, optimal bandwidth) of CO2 emission growth
rates (i.e.,
ln EMit ) of non-Kyoto and Kyoto countries, where t = 0 is the pretreatment (1997 to 2000)
average and t = 1 is the posttreatment (2004 to 2007) average. The graph also shows the mean emission
growth rate in each subpopulation.
22
23
Method
Kyoto (0,1)
FE-OLS
(1)
FE-OLS
(2)
FE-IV
(3)
Long
FE-OLS
(4)
Long
FE-IV
(5)
0.17***
(0.03)
0.06**
(0.02)
0.01**
(0.01)
0.65
(0.55)
0.00
(0.01)
1.10***
(0.24)
0.12
(0.10)
0.09*
(0.05)
0.10
(0.10)
0.18**
(0.07)
0.15**
(0.06)
0.01
(0.07)
0.01
(0.04)
0.00
(0.00)
1,418
133
0.10**
(0.05)
0.02***
(0.00)
0.62
(0.50)
0.00
(0.01)
0.95***
(0.26)
0.12
(0.09)
0.10**
(0.04)
0.10
(0.09)
0.18***
(0.07)
0.15**
(0.06)
0.02
(0.07)
0.00
(0.04)
0.00
(0.00)
1,418
133
0.09**
(0.04)
0.03***
(0.01)
1.13**
(0.54)
0.02
(0.01)
1.12***
(0.26)
0.25***
(0.09)
0.26***
(0.08)
0.26*
(0.15)
0.23**
(0.10)
0.29**
(0.15)
0.16
(0.12)
0.04
(0.05)
0.01
(0.01)
266
133
0.12*
(0.07)
0.03***
(0.01)
1.02*
(0.56)
0.01
(0.02)
1.05***
(0.29)
0.25***
(0.09)
0.27***
(0.09)
0.26*
(0.15)
0.21*
(0.11)
0.28*
(0.15)
0.16
(0.12)
0.03
(0.06)
0.01
(0.01)
266
133
1,456
133
0.28
0.44
19.09
0.27
11.99
0.48
13.85
17.44
0.43
0.71
37.70
0.49
21.40
23.09
Note: Standard errors in parentheses adjusted for within-group clustering and heteroskedasticity. *P <
0.1; **P < 0.05; ***P < 0.01. Instruments for Kyoto variable: ICC membership and its spatial lag. All
regressions use a comprehensive set of year dummies (not reported).
26
We do not report adjusted R2 for IV regressions because they have no statistical meaning (e.g.,
Wooldridge, 2012, p. 523).
27 All appendices are available at the end of this article as it appears in JPAM online. Go to the
publishers Web site and use the search engine to locate the article at http://www3.interscience.
wiley.com/cgi-bin/jhome/34787.
28 LIML is a linear combination of the FE and the FE-IV estimator (with the weights depending on the
data), and the weights happen to be such that they (approximately) eliminate the FE-IV bias resulting
from weak instruments. In theory, LIML has better small sample properties, so that differences in
estimates would reflect the presence of a bias.
0.06
(0.04)
518
133
(B) Alternative
IV strategy
LIML
(B1)
GMM
(B2)
(C) Kyoto
definition
Stringency
(C)
0.10**
(0.05)
1,418
133
0.44
19.09
0.10**
(0.04)
1,418
133
0.44
19.09
0.05**
(0.02)
1,418
133
0.41
21.94
W/o EIT
(D1)
W/o OPEC
(D2)
Rich only
(D3)
Large only
(D4)
0.10*
(0.05)
1,161
109
0.36
17.08
0.10**
(0.05)
1,354
127
0.67
18.18
0.11***
(0.04)
720
67
0.10
8.30
0.09
(0.06)
1,007
94
0.07
16.28
(E) Alternative
treatment window
Kyoto (0,1)
No. of observations
No. of countries
Over-ID test (P-value)
Weak-ID test (F-stat)
(F) GHGs
Narrow
(E1)
Broad
(E2)
2000,
2005
(F1)
2000,
2008
(F2)
0.12**
(0.06)
266
133
0.85
34.06
0.17**
(0.07)
258
129
0.82
37.47
0.10***
(0.03)
218
109
0.48
10.33
0.27***
(0.10)
200
100
0.41
11.59
Note: Dependent variable is ln CO2 emissions, in Panel F, ln GHG emission. FE estimation. In Panels
E and F, long FE estimation. Standard errors in parentheses adjusted for within-group clustering and
heteroskedasticity. *P < 0.1; **P < 0.05; ***P < 0.01. Panel A imposes placebo treatment in 1999 for the
prepolicy period 1997 to 2000. Panels B to F: instruments for Kyoto: ICC membership and its spatial lag.
All regressions use year dummies and additional controls: ln GDP; ln GDP, squared; ln population; ln
IEAs; Chief executive party orientation; openness; WTO dummy; polity. Full results in Appendix Table
A2.29
cap that does not bind (as of 2007), and a value of 2 if it has a binding cap. The
IV strategy continues to work nicely; the Kyoto effect is statistically significant and
quantitatively comparable to the baseline estimations.30
29
All appendices are available at the end of this article as it appears in JPAM online. Go to the publishers Web site and use the search engine to locate the article at http://www3.interscience.wiley.com/
cgi-bin/jhome/34787.
30 Note that the support of the stringency variable is in [0,2] instead of in [0,1].
31
The definition of EIT follows the IMF (2000). The excluded countries are Albania, Armenia, Azerbaijan,
Belarus, Bulgaria, Croatia, Czech Republic, Estonia, FYR of Macedonia, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Poland, Republic of Moldova, Romania, Slovak Republic, Slovenia, Tajikistan,
Turkmenistan, Ukraine, and Uzbekistan. We do not treat Germany as a transition country.
32 The F-stat of 8.3 in column (D3) signals a possible weak instrument problem. However, LIML estimation, as suggested by Stock and Yogo (2005) for such cases, shows that the point estimate of 0.11 is
robust and statistically significant at the 5 percent level. The maximal LIML size bias is 15 percent.
33 This finding may reflect treatment heterogeneity: If small countries are more likely to accept ambitious
binding commitments or are more willing to effectively implement them, their presence in the sample is
necessary for a negative average treatment effect.
34 Kyoto countries that ratified after 2002 are still considered to be treated. The only exceptions being
Australia and Croatia, which ratified Kyoto in late 2007.
35 Only Belarus ratification lies outside the broad treatment window.
36 Note, however, that the sample shrinks to 109 and 100 countries due to data availability.
37 Due to the short time horizon of our analysis, there is a natural limit to the number of lags one can
meaningfully include. We experimented with several combinations. The fourth lag is not statistically
significant and lowered the F-statistic of a test of joint significance of Kyoto and its time lags. Also note
that it is not possible to instrument these lags, so estimates are to be interpreted with care.
Dep. var.
Kyoto (0,1)
Kyoto1 (0,1)
Kyoto2 (0,1)
Kyoto3 (0,1)
No. of
observations
No. of countries
Adj. R2
F-stat
F-test joint
significance
(P-value)
(1)
CO2
emission
(2)
Renewables
(3)
Fossils
0.03*
(0.02)
0.00
(0.01)
0.02
(0.01)
0.04***
(0.02)
0.80**
(0.39)
0.42**
(0.21)
0.26
(0.20)
0.57
(0.36)
1,418
133
0.48
13.28
0.038
Price of
(4)
Coal
(5)
New
Clean
(6)
Diesel
(7)
Gasoline
(8)
Energy
(9)
Electricity
0.56
(0.53)
0.42
(0.39)
0.28
(0.46)
0.94**
(0.45)
0.03
(0.67)
0.80
(0.72)
0.28
(0.60)
0.93*
(0.53)
0.50***
(0.18)
0.07
(0.24)
0.57**
(0.24)
1.23***
(0.31)
0.01
(0.03)
0.05
(0.10)
0.14
(0.10)
0.03
(0.03)
0.05
(0.03)
0.05
(0.07)
0.14**
(0.07)
0.05
(0.03)
0.03*
(0.01)
0.00
(0.01)
0.02**
(0.01)
0.05***
(0.01)
0.03
(0.02)
0.01
(0.01)
0.01
(0.01)
0.03*
(0.02)
1,180
1,180
1,180
1,180
610
610
1,180
1,169
110
0.29
6.11
0.173
110
0.17
5.74
0.127
110
0.21
2.73
0.000
129
0.79
57.50
0.000
129
0.70
51.12
0.000
110
0.39
7.48
0.000
109
0.52
14.25
0.405
110
0.06
1.86
0.067
Note: FE estimations with time lags. Standard errors in parentheses adjusted for within-group clustering and heteroskedasticity. *P < 0.1; **P < 0.05; ***P < 0.01. All regressions use year dummies and additional controls: ln GDP; ln
GDP, squared; ln population; ln share of manufacturing/services/agriculture; ln IEAs; Chief executive party orientation;
openness; WTO dummy; polity (not reported). Full results in Appendix Table A3.38
Summarizing, we employ our instruments in a second-stage regression that estimates the effect of Kyoto commitment on CO2 emissions. Our instruments perform
very well: The weak identification test yields an F-statistic of at least 16.28 and often
substantially higher; the overidentification tests are easily passed. The effect of Kyoto ratification on emissions is robustly negative over all econometric models and
survives a host of robustness checks. The effect is economically substantial: Kyoto
has caused CO2 emissions to fall relative to the counterfactual by about 10 percent.
However, the estimates are not always very precise (though typically different from
zero at the 10 percent to 5 percent level). We find a similar effect for GHG emissions.
CHANNELS: FUEL MIX, FUEL PRICES, AND ELECTRICITY USE
Is a negative relationship between Kyoto commitment and CO2 emissions plausible?
The Kyoto Protocol impacts emissions through domestic policies and measures
such as cap-and-trade systems for GHGs (in the EU, Norway, Switzerland, and
New Zealand), changes in fuel taxes, feed-in tariffs for alternative energy sources,
regulations (of, e.g., cars fuel economy, insulation requirements for houses, low
energy bulbs), and energy efficiency and mandatory performance standards for
electrical products. In the following subsections, we present evidence that sheds
light on the channels through which Kyoto may have led to lower emissions. We
focus on energy and electricity mix, fuel prices, as well as energy and electricity use.
The selection of channels is certainly exploratory, and it is mostly driven by data
availability.39
38
All appendices are available at the end of this article as it appears in JPAM online. Go to the publishers Web site and use the search engine to locate the article at http://www3.interscience.wiley.com/
cgi-bin/jhome/34787.
39 For example, there is not sufficient data to investigate the channels of regulation or labeling.
.2
1.5
.15
.2
.1
10
10
.5
10 5
10
15
20
10
10
20
10
Diesel price
Gasoline price
Ln energy/capita
Ln electricity/capita
40
60
US cents
80
.03
.01
50
50
100
US cents
20
.02
.02
.01
.01
.02
.02
.04
.03
.04
.05
percentage points
.06
percentage points
.04
percentage points
percentage points
.03
20
.05
0
.05
.1
.1
.2
.15
.3
Coal (##)
.4
Renewables (#)
40
20
Kyoto
20
40
50
50
NonKyoto
Note: The diagrams show kernel density plots (Epanechnikov, optimal bandwidth) of changes in respective variables (i.e.,
Yit ) of non-Kyoto and Kyoto countries, where t = 0 is the pretreatment (1997 to
2000) average and t = 1 is the posttreatment (2004 to 2007) average. New clean refers to wind and solar
energy. #: Share in total energy use. ##: Share in electricity production.
Figure 2. Changes in Energy Mix, Fuel Prices, Energy Use, and Kyoto Commitment.
However, changes in the energy mix could indicate that Kyoto works through
feed-in tariffs or other ways to promote green energy (e.g., R&D subsidies for green
energy sources). Likewise, changes in fuel prices will pick up effects from fuel
taxes. On the other hand, changes in energy and electricity use are not directly
related to one specific policy measure, but rather reflect changes along all policy
dimensions combined. But these outcome variables help to disentangle technique
and composition from scale effects (see, e.g., Copeland & Taylor, 2003). Hence,
we can learn whether emission mitigation results from a switch toward less CO2 intensive production technologies (technique effect) and changes in the composition
of produced goods (composition effect)both resulting in a lower energy intensity
or whether adjustments occur along the scale of production (scale effect).
Evidence for effects along these channels would support our findings in the section
on Kyotos effect on emissions. As discussed for the case of CO2 emissions, the same
endogeneity issues arise. For example, countries with a higher share of wind, solar,
or nuclear energy may find it easier to commit to Kyoto. Therefore, we provide FE
OLS and IV results based on the same models and instrumentation strategy.
Energy Mix
We start with a graphical inspection of the effect of Kyoto commitments on countries energy mix. The upper leftmost diagram in Figure 2 shows a kernel density plot
Journal of Policy Analysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management
40
41
42
This energy source comprises solid and liquid biomass, biogas, industrial, and municipal waste.
Fossil fuel comprises coal, oil, petroleum, and natural gas products.
To save space, Table 6 only reports Kyoto estimates and first- and second-stage diagnostics. Full
results are delegated to Appendix Table A4. All appendices are available at the end of this article as it
appears in JPAM online. Go to the publishers Web site and use the search engine to locate the article at
http://www3.interscience.wiley.com/cgi-bin/jhome/34787.
Dep. var.
Renewables
Fossil fuel
Alternative
energy
(A1)
(A2)
(A3)
(A4)
(B1)
(B2)
(B3)
(B4)
FE-OLS FE-IV FE-OLS FE-IV FE-OLS FE-IV FE-OLS FE-IV
Method
Kyoto (0,1)
No. of observations
No. of countries
Sheas partial R2
Over-ID test (P-value)
Weak-ID test (F-stat)
Adj. R2
Panel
1.38**
(0.56)
1,180
110
2.41***
(0.93)
1,180
110
0.26
0.68
18.80
0.29
0.67
(0.63)
1,180
110
2.46**
(1.16)
1,180
110
0.26
0.47
18.80
0.17
Diesel fuel
1.43
(1.76)
1,180
110
0.26
0.60
18.80
0.06
Dep. var.
0.12
(0.93)
1,180
110
Gasoline
1.07***
(0.27)
1,180
110
1.66***
(0.56)
1,180
110
0.26
0.63
18.80
0.17
Electricity
(C1)
(C2)
(C3)
(C4)
(D1)
(D2)
(D3)
(D4)
FE-OLS FE-IV FE-OLS FE-IV FE-OLS FE-IV FE-OLS FE-IV
Method
Kyoto (0,1)
No. of observations
No. of countries
Sheas partial R2
Over-ID test (P-value)
Weak-ID test (F-stat)
Adj. R2
0.10**
(0.04)
608
127
0.77
0.22***
(0.05)
608
127
0.28
0.61
20.30
0.13***
(0.04)
608
127
0.68
0.25***
(0.05)
608
127
0.28
0.32
20.30
0.05** 0.05*
(0.02) (0.03)
1,180
1,180
110
110
0.26
0.36
18.80
0.38
0.04
(0.03)
1,169
109
0.08**
(0.04)
1,169
109
0.26
0.21
18.84
0.52
Note: Standard errors in parentheses adjusted for within-group clustering and heteroskedasticity.
*P < 0.1; **P < 0.05; ***P < 0.01. Instruments for Kyoto: ICC membership and its spatial lag. All regressions use year dummies and additional controls: ln GDP; ln GDP, squared; ln population; ln share of
manufacturing/services/agriculture; ln IEAs; Chief executive party orientation; openness; WTO dummy;
polity (not reported). Full results in Appendix Table A4.43
Panel B of Table 6 shows regression results. Columns B1 and B2 analyze the coal
share in electricity production. Controlling for a host of variables and applying our
IV strategy, there is no statistical evidence in favor of a negative effect of Kyoto on
the coal share. This is as suggested by our graphical analysis. One may view this as a
surprising result since coal is the most CO2 intensive fuel for electricity production.
However, it seems that countries have preferred to cut back on the use of other
fossil fuels, especially petrol, which is more likely to be imported, therefore meeting
less local political resistance. Columns B3 and B4 study the share of alternative
energy sources. The IV regression suggests that Kyoto commitments have increased
that share by about 1.66 percentage points. It appears that CO2 -free energy sources
43
All appendices are available at the end of this article as it appears in JPAM online. Go to the publishers Web site and use the search engine to locate the article at http://www3.interscience.wiley.com/
cgi-bin/jhome/34787.
Fuel Prices
We continue with the fuel price channel. The two lower left panels of Figure 2 plot
kernel densities of the (absolute) changes in fuel prices, expressed in U.S. cents per
liter, across the groups of committed and noncommitted countries. Between the
pretreatment and posttreatment period, the diesel price has increased by about 30
cents in the group of noncommitted countries and by 49 in committed countries.
The differential increase across the two groups is 18 cents. It is significant at the
1 percent level. A similar picture emerges when looking at gasoline. The average
increase in noncommitted countries was 29 cents and in committed countries 45.
The difference, 16 cents, is again statistically significant at the 1 percent level. Note
that comparing changes over time between Kyoto and non-Kyoto countries, we are
effectively controlling for changes in the world market prices for these fuels.
Panel C in Table 6 confirms these patterns. Column C1 reports the uninstrumented
FE estimator. It suggests that Kyoto status is positively associated to the price of
diesel fuel. With an adjusted R2 of 77 percent, the specification is surprisingly successful in predicting the diesel price. Instrumenting Kyoto commitment leaves the
controls virtually unchanged, but the Kyoto effect more than doubles to 22 cents per
liter. The IV strategy turns out to work reasonably well: The weak identification test
yields an F-statistic of 20.3, and the overidentification test does not reject instrument
validity. The results for the price of gasoline (columns C3 and C4) look very similar.
Again, the IV estimation yields a Kyoto point estimate that is about double the OLS
estimate. In summary, Kyoto commitment has increased fuel prices considerably.
Given the design of our econometric exercise where year dummies control for the
world price of the fuel and country FE control for endowments and proximity to
these endowments, we may interpret these findings as indirect evidence that Kyoto
commitment has led to higher fuel taxes.
44
The overidentification test signals endogenous instruments in the long FE estimation for diesel prices,
the FE model on rich countries only for gasoline prices, and the specification without EITs for per capita
electricity use.
Long FE
W/O EIT
Rich only
Long FE
W/O EIT
Rich only
Fossil fuel
(A1)
(A2)
(A3)
(A4)
(A5)
(A6)
3.43***
(1.14)
220
0.32
36.49
3.44***
(1.29)
923
0.53
15.81
2.05**
(0.85)
701
0.72
8.39
3.62***
(1.37)
220
0.18
36.49
2.80**
(1.29)
923
0.50
15.81
0.58
(1.13)
701
0.58
8.39
Coal
(B1)
(B2)
(B3)
(B4)
(B5)
(B6)
3.06***
(0.58)
220
0.60
36.49
2.06***
(0.56)
923
0.55
15.81
1.69***
(0.54)
701
0.30
8.39
0.96
(1.41)
220
0.62
36.49
1.71
(2.29)
923
0.47
15.81
0.74
(1.30)
701
0.92
8.39
Gasoline
(C1)
(C2)
(C3)
(C4)
(C5)
(C6)
0.49***
(0.06)
252
0.04
39.06
0.12**
(0.05)
498
0.67
18.86
0.31***
(0.08)
316
0.39
8.83
0.46***
(0.06)
252
0.47
39.06
0.17***
(0.06)
498
0.34
18.86
0.32***
(0.09)
316
0.04
8.83
Electricity
(D1)
(D2)
(D3)
(D4)
(D5)
(D6)
0.06
(0.04)
220
0.67
36.49
0.07*
(0.04)
923
0.31
15.81
0.06*
(0.03)
701
0.42
8.39
0.02
(0.06)
218
0.92
36.30
0.13***
(0.04)
912
0.03
16.23
0.06
(0.04)
701
0.74
8.39
Note: Standard errors in parentheses adjusted for within-group clustering and heteroskedasticity. *P <
0.1; **P < 0.05; ***P < 0.01. Instruments for Kyoto: ICC membership and its spatial lag. All regressions
use year dummies and additional controls as in Table 6. Full results in Appendix Table A5.45
45
All appendices are available at the end of this article as it appears in JPAM online. Go to the
publishers Web site and use the search engine to locate the article at http://www3.interscience.
wiley.com/cgi-bin/jhome/34787.
REFERENCES
Aakvik, A., & Tjtta, S. (2011). Do collective actions clear common air? The effect of international environmental protocols on sulphur emissions. European Journal of Political
Economy, 27, 343351.
Aichele, R., & Felbermayr, G. (2012). Kyoto and the carbon footprint of nations. Journal of
Environmental Economics and Management, 63, 336354.
Andreoni, J., & Levinson, A. (2001). The simple analytics of the environmental Kuznets curve.
Journal of Public Economics, 80, 269286.
Andreoni, J., & McGuire, M. C. (1993). Identifying the free riders: A simple algorithm for
determining who will contribute to a public good. Journal of Public Economics, 51, 447
454.
Azomahou, T., Laisney, F., & Nguyen-Van, P. (2006). Economic development and CO2 emissions: A nonparametric panel approach. Journal of Public Economics, 90, 13471363.
Beron, K. J., Murdoch, J. C., & Vijverberg, W. P. M. (2003). Why cooperate? Public goods,
economic power, and the Montreal protocol. Review of Economics and Statistics, 85, 286
297.
Bertrand, M., Duflo, E., & Mullainathan, S. (2004). How much should we trust differencesin-differences estimates? Quarterly Journal of Economics, 119, 249275.
Brock, W. A., & Taylor, M. S. (2010). The green Solow model. Journal of Economic Growth,
15, 127153.
Carraro, C., & Siniscalco, D. (1998). International institutions and environmental policy: International environmental agreements: Incentives and political economy. European Economic Review, 42, 561572.
Cole, M. A., & Neumayer, E. (2004). Examining the impact of demographic factors on air
pollution. Population & Environment, 26, 521.
Congleton, R. (2001). Governing the global environmental commons: The political economy of
international environmental treaties and institutions. In G. Schulze & H. Ursprung (Eds.),
Globalization and the environment (pp. 241263). New York, NY: Oxford University Press.
Copeland, B. R., & Taylor, M. S. (2003). Trade and the environment: Theory and evidence.
Princeton and Oxford: Princeton University Press.
Dinda, S. (2004). Environmental Kuznets curve hypothesis: A survey. Ecological Economics,
49, 431455.
Finus, M., & Tjtta, S. (2003). The Oslo protocol on sulfur reduction: The great leap forward?
Journal of Public Economics, 87, 20312048.
Journal of Policy Analysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management
ISO3
Country
AGO
ALB
ARE
ARG
ARM
AUS
AUT
AZE
BDI
BEL
BEN
BFA
BGD
BGR
BLR
BOL
BRA
BTN
BWA
CAN
CHL
CHN
CIV
CMR
COG
COL
COM
CRI
CUB
CYP
CZE
DEU
DJI
DNK
DOM
DZA
EGY
ERI
ESP
EST
ETH
FIN
FJI
FRA
GAB
GIN
GMB
GNQ
Angola
Albania
United Arab Emirates
Argentina
Armenia
Australia
Austria
Azerbaijan
Burundi
Belgium
Benin
Burkina Faso
Bangladesh
Bulgaria
Belarus
Bolivia
Brazil
Bhutan
Botswana
Canada
Chile
China
Cote dIvoire
Cameroon
Congo, Rep.
Colombia
Comoros
Costa Rica
Cuba
Cyprus
Czech Republic
Germany
Djibouti
Denmark
Dominican Republic
Algeria
Egypt, Arab Rep.
Eritrea
Spain
Estonia
Ethiopia
Finland
Fiji
France
Gabon
Guinea
The Gambia
Equatorial Guinea
Sample
Emission growth Kyoto ICC
(in percent)
(year) (year) Rich Large OPEC EIT
91
66
30
18
28
8
13
11
55
8
81
36
44
2
10
19
9
32
23
11
12
55
3
33
27
3
40
32
3
16
1
6
12
5
6
22
30
12
23
9
16
11
72
1
10
7
34
261
n.a.
n.a.
n.a.
n.a.
n.a.
2007
2002
n.a.
n.a.
2002
n.a.
n.a.
n.a.
2002
2005
n.a.
n.a.
n.a.
n.a.
2002
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2001
2002
n.a.
2002
n.a.
n.a.
n.a.
n.a.
2002
2002
n.a.
2002
n.a.
2002
n.a.
n.a.
n.a.
n.a.
n.a.
2003
n.a.
2001
n.a.
2002
2000
n.a.
2004
2000
2002
2004
n.a.
2002
n.a.
2002
2002
n.a.
2000
2000
n.a.
n.a.
n.a.
n.a.
2004
2002
2006
2001
n.a.
2002
n.a.
2000
2002
2001
2005
n.a.
n.a.
n.a.
2000
2002
n.a.
2000
1999
2000
2000
2003
2002
n.a.
0
0
1
1
1
1
1
1
0
1
0
0
0
1
1
0
1
0
1
1
1
1
0
0
0
1
0
1
1
1
1
1
0
1
1
0
0
0
1
1
0
1
0
1
1
0
0
1
1
0
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
1
1
0
1
0
0
1
0
1
1
0
1
1
1
1
1
1
0
1
1
0
1
0
1
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
1
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
ISO3
Country
Emission growth
(in percent)
Kyoto
(year)
ICC
(year)
Rich
Large
OPEC
EIT
GRC
GTM
GUY
HND
HRV
HUN
IDN
IND
IRL
IRN
ITA
JAM
JOR
JPN
KAZ
KEN
KGZ
KHM
KOR
LAO
LBR
LKA
LTU
LVA
MAR
MDA
MDG
MEX
MKD
MLI
MNG
MOZ
MRT
MUS
MWI
MYS
NAM
NIC
NLD
NOR
NPL
NZL
PAK
PRT
PRY
ROM
RWA
SAU
SDN
SEN
Greece
Guatemala
Guyana
Honduras
Croatia
Hungary
Indonesia
India
Ireland
Iran, Islamic Rep.
Italy
Jamaica
Jordan
Japan
Kazakhstan
Kenya
Kyrgyz Republic
Cambodia
Korea, Rep.
Lao PDR
Liberia
Sri Lanka
Lithuania
Latvia
Morocco
Moldova
Madagascar
Mexico
Macedonia, FYR
Mali
Mongolia
Mozambique
Mauritania
Mauritius
Malawi
Malaysia
Namibia
Nicaragua
Netherlands
Norway
Nepal
New Zealand
Pakistan
Portugal
Paraguay
Romania
Rwanda
Saudi Arabia
Sudan
Senegal
11
32
9
51
15
4
36
28
11
37
5
15
34
1
43
15
9
61
16
50
59
34
0
2
29
11
7
14
5
7
20
56
1
41
14
43
40
19
1
22
11
6
36
2
5
4
5
46
75
37
2002
n.a.
n.a.
n.a.
2007
2002
n.a.
n.a.
2002
n.a.
2002
n.a.
n.a.
2002
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2003
2002
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2002
2002
n.a.
2002
n.a.
2002
n.a.
2001
n.a.
n.a.
n.a.
n.a.
2002
n.a.
2004
2002
2001
2001
n.a.
n.a.
2002
n.a.
1999
n.a.
2002
2007
n.a.
2005
n.a.
2002
2002
n.a.
2006
n.a.
2003
2002
n.a.
n.a.
n.a.
2005
2002
2000
2002
n.a.
n.a.
2002
2002
n.a.
2002
n.a.
2001
2000
n.a.
2000
n.a.
2002
2001
2002
n.a.
n.a.
n.a.
1999
1
0
0
0
1
1
0
0
1
1
1
1
0
1
1
0
0
0
1
0
0
0
1
1
0
0
0
1
0
0
0
0
0
1
0
1
0
0
1
1
0
1
0
1
0
1
0
1
0
0
1
1
0
1
0
1
1
1
0
1
1
0
1
1
1
1
1
1
1
1
0
1
0
0
1
0
1
1
0
1
0
1
0
0
1
1
0
1
1
0
1
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
1
1
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
ISO3
Country
SGP
SLB
SLV
SVK
SVN
SWE
SWZ
SYR
TCD
TGO
THA
TJK
TKM
TTO
TUN
TUR
TZA
UGA
UKR
URY
USA
UZB
VEN
VNM
ZAF
ZMB
ZWE
Singapore
Solomon Islands
El Salvador
Slovak Republic
Slovenia
Sweden
Swaziland
Syrian Arab Rep.
Chad
Togo
Thailand
Tajikistan
Turkmenistan
Trinidad and Tobago
Tunisia
Turkey
Tanzania
Uganda
Ukraine
Uruguay
United States
Uzbekistan
Venezuela, RB
Vietnam
South Africa
Zambia
Zimbabwe
Sample
Emission growth Kyoto ICC
(in percent)
(year) (year) Rich Large OPEC EIT
7
12
11
5
1
3
16
13
110
4
32
20
30
45
23
21
67
62
2
7
4
1
6
77
11
17
36
n.a.
n.a.
n.a.
2002
2002
2002
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2004
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2002
2001
2001
n.a.
n.a.
2006
n.a.
n.a.
2000
n.a.
1999
n.a.
n.a.
2002
2002
n.a.
2002
n.a.
n.a.
2000
n.a.
2000
2002
n.a.
1
0
0
1
1
1
0
0
0
0
1
0
1
1
1
1
0
0
1
1
1
0
1
0
1
0
0
0
0
1
1
0
1
0
1
1
1
1
1
0
0
1
1
1
1
1
0
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
1
1
0
0
0
0
0
1
0
0
1
0
0
0
0
0
Note: Emission growth is difference between pretreatment and posttreatment (i.e., 1997 to 2000 and
2004 to 2007, respectively) average of log CO2 emissions. Kyoto shows the ratification year of Kyoto
commitment, ICC shows the ratification year of the Rome Statutes of the ICC. EIT is a dummy for
economies in transition.
Due to data availability, the following countries drop out of the sample (compared to the CO2 emissions
data from WDI): Andorra, Antigua and Barbuda, Aruba, Bahamas, Barbados, Belize, Brunei Darussalam, Cape Verde, Central African Republic, Dem. Rep. of the Congo, Dominica, Ecuador, Grenada,
Guinea-Bissau, Haiti, Iceland, Iraq, Israel, Kiribati, Korea Dem. Rep., Kuwait, Lebanon, Liechtenstein,
Luxembourg, Libya, Maldives, Malta, Marshall Islands, Micronesia, Monaco, Montenegro, Myanmar,
Netherlands Antilles, Niger, Nigeria, Oman, Palau, Qatar, Russia, Saint Lucia, Samoa, San Marino, Sao
Tome and Principe, Serbia, Seychelles, Sierra Leone, St. Kitts and Nevis, St. Vincent and the Grenadines,
Suriname, Tonga, Tuvalu, Vanuatu, and Yemen.
Furthermore, the following nonindependent territories are also dropped: American Samoa, Bermuda,
Cayman Islands, Channel Islands, Faroe Islands, French Polynesia, Gibraltar, Greenland, Guam, Hong
Kong, Isle of Man, Kosovo, Macao, Mayotte, New Caledonia, Northern Mariana Islands, Puerto Rico,
Turks and Caicos, Virgin Islands (United States), and West Bank and Gaza.
0.06
(0.04)
0.03
(0.05)
1.37
(1.01)
0.06*
(0.03)
1.34
(1.21)
0.06
(0.07)
0.04
(0.07)
0.11
(0.13)
0.16
(0.15)
0.19
(0.17)
0.11
(0.09)
0.04
(0.08)
0.00
(0.00)
518
133
0.11
(B) Alternative
IV strategy
LIML
(B1)
GMM
(B2)
(C) Kyoto
definition
Stringency
(C)
0.10**
(0.05)
0.02***
(0.00)
0.62
(0.50)
0.00
(0.01)
0.95***
(0.26)
0.12
(0.09)
0.10**
(0.04)
0.10
(0.09)
0.18***
(0.07)
0.15**
(0.06)
0.02
(0.07)
0.00
(0.04)
0.00
(0.00)
1,418
133
0.10**
(0.04)
0.02***
(0.00)
0.58
(0.50)
0.00
(0.01)
0.98***
(0.26)
0.12
(0.09)
0.09**
(0.04)
0.10
(0.09)
0.17**
(0.07)
0.14**
(0.06)
0.02
(0.07)
0.00
(0.04)
0.00
(0.00)
1,418
133
0.05**
(0.02)
0.02***
(0.00)
0.66
(0.51)
0.00
(0.01)
0.99***
(0.25)
0.12
(0.09)
0.09**
(0.04)
0.10
(0.09)
0.17**
(0.07)
0.14**
(0.06)
0.00
(0.07)
0.00
(0.04)
0.00
(0.00)
1,418
133
0.44
19.09
0.44
19.09
0.41
21.94
0.42
0.42
0.42
W/o EIT
(D1)
W/o OPEC
(D2)
Rich only
(D3)
Large only
(D4)
0.11**
(0.05)
0.02***
(0.00)
0.92*
(0.52)
v0.01
(0.01)
0.62**
(0.31)
0.17*
(0.10)
0.12**
(0.05)
0.16
(0.10)
0.14*
(0.08)
0.13**
(0.06)
0.02
(0.08)
0.02
(0.03)
0.00
(0.00)
1,172
110
0.10**
(0.05)
0.02***
(0.00)
0.55
(0.52)
0.00
(0.01)
1.02***
(0.30)
0.12
(0.10)
0.11**
(0.05)
0.07
(0.11)
0.15**
(0.07)
0.14**
(0.06)
0.02
(0.07)
0.00
(0.04)
0.00
(0.00)
1,354
127
0.11***
(0.04)
0.03
(0.02)
0.75
(0.70)
0.00
(0.02)
0.55*
(0.33)
0.32**
(0.15)
0.05
(0.05)
0.03
(0.21)
0.21***
(0.06)
0.06
(0.06)
0.25***
(0.07)
0.02
(0.05)
0.01*
(0.01)
720
67
0.09
(0.06)
0.02***
(0.00)
0.45
(0.58)
0.03*
(0.01)
1.38***
(0.39)
0.03
(0.07)
0.12**
(0.06)
0.14
(0.12)
0.19**
(0.08)
0.13*
(0.07)
0.09
(0.08)
0.01
(0.04)
0.00
(0.00)
1,007
94
0.36
21.37
0.67
18.18
0.10
8.30
0.07
16.28
0.46
0.41
0.43
0.51
Kyoto (0,1)
Kyoto, spatial lag
ln GDP
ln GDP, squared
ln population
ln manufacturinga
ln agriculturea
ln servicesa
ln stock of other MEA
Government orientation (0.1, 0.2, 0.3)
Openness, (Exp + Imp)/GDP
WTO (0,1)
Polity (1 to 1)
No. of observations
No. of countries
First-stage diagnostics
Over-ID test (P-value)
Weak-ID test (F-stat)
Second-stage diagnostics
Adj. R2
(F) GHGs
Narrow
(E1)
Broad
(E2)
2000, 2005
(F1)
2000, 2008
(F2)
0.12**
(0.06)
0.04***
(0.01)
0.85
(0.57)
0.01
(0.02)
1.01***
(0.30)
0.15
(0.11)
0.22***
(0.08)
0.24
(0.16)
0.17*
(0.10)
0.19
(0.16)
0.12
(0.13)
0.03
(0.06)
0.01
(0.01)
266
133
0.17**
(0.07)
0.04***
(0.01)
0.63
(0.51)
0.00
(0.01)
1.04***
(0.26)
0.11
(0.08)
0.18**
(0.09)
0.09
(0.14)
0.24**
(0.10)
0.26*
(0.15)
0.06
(0.13)
0.06
(0.06)
0.01
(0.01)
258
129
0.10***
(0.03)
0.00*
(0.00)
0.01
(0.31)
0.02
(0.02)
0.15
(0.30)
0.06
(0.06)
0.00
(0.05)
0.13
(0.11)
0.05
(0.05)
0.00***
(0.00)
0.00
(0.00)
0.07
(0.05)
0.00
(0.00)
218
109
0.27***
(0.10)
0.00**
(0.00)
1.46
(1.06)
0.10*
(0.06)
0.36
(0.46)
0.03
(0.08)
0.07
(0.07)
0.35
(0.33)
0.04
(0.13)
0.00
(0.00)
0.00
(0.00)
0.13
(0.11)
0.00
(0.01)
200
100
0.85
34.06
0.82
37.47
0.48
10.33
0.41
11.59
0.38
0.32
0.14
0.25
Note: Dependent variable is ln CO2 emissions, in Panel F, ln GHG emissions. Long FE estimation. All
regressions use period dummies and constant (not shown). Standard errors in parentheses adjusted for
within-group clustering and heteroskedasticity. *P < 0.1; **P < 0.05; ***P < 0.01. Instruments for Kyoto:
ICC membership and its spatial lag.
a In
percent of GDP.
1,418
0.48
13.28
0.03
(0.02)
0.00
(0.01)
0.02
(0.01)
0.04***
(0.02)
0.01**
(0.01)
0.65
(0.54)
0.00
(0.01)
1.02***
(0.25)
0.13
(0.10)
0.10**
(0.05)
0.10
(0.10)
0.18**
(0.07)
0.15**
(0.06)
0.01
(0.08)
0.00
(0.04)
0.00
(0.00)
CO2 emission
1,180
0.29
6.11
0.80
(0.39)
0.42**
(0.21)
0.26
(0.20)
0.57
(0.36)
0.59***
(0.18)
1.02
(9.94)
0.14
(0.28)
15.25*
(7.70)
1.49
(1.21)
1.74
(1.28)
0.82
(2.54)
1.45
(1.14)
0.04
(1.72)
0.12
(1.30)
0.42
(0.72)
0.11*
(0.06)
**
Renewables
1,180
0.17
5.74
0.56
(0.53)
0.42
(0.39)
0.28
(0.46)
0.94**
(0.45)
0.53**
(0.23)
1.25
(10.95)
0.24
(0.30)
15.42**
(7.66)
0.14
(1.41)
2.14
(1.36)
1.73
(2.74)
2.03
(1.44)
3.60*
(1.94)
0.14
(1.33)
0.21
(1.03)
0.07
(0.07)
Fossils
Share of
1,180
0.06
1.86
0.03
(0.67)
0.80
(0.72)
0.28
(0.60)
0.93*
(0.53)
0.21
(0.19)
7.65
(10.58)
0.35
(0.29)
6.17
(6.19)
2.53*
(1.31)
0.64
(1.13)
1.12
(1.17)
1.44
(1.35)
0.18
(2.53)
1.91
(1.31)
0.13
(0.91)
0.17
(0.12)
Coal
1,180
0.21
2.73
0.50
(0.18)
0.07
(0.24)
0.57**
(0.24)
1.23***
(0.31)
0.01
(0.06)
0.12
(3.35)
0.04
(0.10)
2.67
(1.88)
0.14
(0.49)
0.01
(0.33)
0.08
(0.51)
0.62
(0.47)
0.10
(1.22)
0.25
(0.35)
0.24
(0.19)
0.06
(0.06)
***
Clean
610
0.79
57.50
0.01
(0.03)
0.05
(0.10)
0.14
(0.10)
0.03
(0.03)
0.01*
(0.01)
1.22**
(0.51)
0.03**
(0.01)
0.19
(0.30)
0.03
(0.05)
0.05
(0.06)
0.21**
(0.09)
0.08
(0.07)
0.15
(0.13)
0.01
(0.07)
0.02
(0.04)
0.00
(0.00)
Diesel
610
0.70
51.12
0.05
(0.03)
0.05
(0.07)
0.14**
(0.07)
0.05
(0.03)
0.01
(0.01)
0.55
(0.61)
0.01
(0.02)
0.35
(0.36)
0.02
(0.07)
0.05
(0.09)
0.27**
(0.11)
0.11
(0.15)
0.19
(0.15)
0.08
(0.10)
0.02
(0.05)
0.00
(0.00)
Gasoline
Price of
1,180
0.39
7.48
0.03
(0.01)
0.00
(0.01)
0.02**
(0.01)
0.05***
(0.01)
0.00
(0.00)
0.57*
(0.32)
0.02**
(0.01)
0.43**
(0.18)
0.04
(0.04)
0.06*
(0.04)
0.12*
(0.07)
0.02
(0.05)
0.01
(0.05)
0.06*
(0.03)
0.02
(0.02)
0.00
(0.00)
*
Energy
1,169
0.52
14.25
0.03
(0.02)
0.01
(0.01)
0.01
(0.01)
0.03*
(0.02)
0.00
(0.01)
0.04
(1.05)
0.01
(0.03)
0.51
(0.35)
0.20*
(0.12)
0.01
(0.05)
0.08
(0.08)
0.16
(0.12)
0.10
(0.08)
0.12
(0.10)
0.08*
(0.05)
0.00
(0.00)
Electricity
Note: FE estimations with time lags. Standard errors in parentheses adjusted for within-group clustering and heteroskedasticity. *P < 0.1; **P < 0.05; ***P < 0.01. All regressions use
year dummies and constant (not shown).
Observations
Adj. R2
F-statistic
Polity (1 to 1)
WTO (0,1)
Government orientation
ln services
ln agriculture
ln manufacturing
ln population
ln GDP, squared
ln GDP
Kyoto3 (0,1)
Kyoto2 (0,1)
Kyoto1 (0,1)
Kyoto (0,1)
Dep. var.
Polity (1 to 1)
WTO (0,1)
Openness
Government orientation
ln MEA stock
ln services
ln agriculture
ln manufacturing
ln population
ln GDP, squared
ln GDP
Kyoto (0,1)
Method
Dep. var.
Panel
2.41***
(0.93)
0.62***
(0.15)
0.02
(9.24)
0.17
(0.26)
13.31*
(7.80)
1.35
(1.14)
1.76
(1.17)
0.74
(2.40)
1.40
(1.06)
0.01
(1.61)
0.07
(1.17)
0.36
(0.67)
0.11*
(0.06)
1,180 (110)
0.68
18.80
0.21
1.38**
(0.56)
0.58***
(0.18)
1.16
(9.99)
0.15
(0.28)
16.89**
(7.45)
1.34
(1.25)
1.58
(1.28)
0.77
(2.53)
1.47
(1.16)
0.01
(1.71)
0.18
(1.29)
0.53
(0.71)
0.11*
(0.06)
1,180 (110)
0.29
FE-IV
FE-OLS
Renewables
2.46**
(1.16)
0.61***
(0.16)
3.15
(10.32)
0.29
(0.28)
10.58
(8.05)
0.26
(1.30)
2.31*
(1.24)
1.64
(2.62)
1.94
(1.37)
3.59**
(1.82)
0.30
(1.25)
0.01
(0.99)
0.07
(0.07)
1,180 (110)
0.47
18.80
0.06
0.67
(0.63)
0.52**
(0.22)
1.18
(11.02)
0.25
(0.30)
16.78**
(7.46)
0.28
(1.45)
2.00
(1.35)
1.69
(2.74)
2.05
(1.45)
3.60*
(1.91)
0.14
(1.33)
0.31
(1.02)
0.07
(0.07)
1,180 (110)
0.17
FE-IV
FE-OLS
Fossil fuel
Table A4. Through which channels does Kyoto operate? Full results.
0.06
0.12
(0.93)
0.22
(0.19)
7.62
(10.58)
0.36
(0.29)
7.14
(5.58)
2.41*
(1.30)
0.76
(1.15)
1.14
(1.16)
1.46
(1.35)
0.17
(2.57)
1.94
(1.31)
0.06
(0.94)
0.17
(0.12)
1,180 (110)
FE-OLS
FE-IV
1.43
(1.76)
0.14
(0.11)
9.33
(10.36)
0.40
(0.28)
1.77
(7.93)
2.43**
(1.21)
0.49
(1.12)
1.19
(1.13)
1.36
(1.29)
0.16
(2.43)
2.32*
(1.40)
0.33
(1.00)
0.17
(0.11)
1,180 (110)
0.60
18.80
0.05
Coal
0.17
1.07***
(0.27)
0.03
(0.06)
0.30
(3.40)
0.05
(0.10)
0.12
(1.91)
0.39
(0.51)
0.26
(0.37)
0.15
(0.52)
0.57
(0.47)
0.14
(1.33)
0.19
(0.39)
0.42**
(0.20)
0.06
(0.06)
1,180 (110)
FE-OLS
1.66***
(0.56)
0.00
(0.03)
0.35
(3.37)
0.06
(0.10)
2.18
(2.59)
0.39
(0.46)
0.16
(0.36)
0.17
(0.49)
0.61
(0.43)
0.13
(1.24)
0.34
(0.42)
0.32
(0.22)
0.06
(0.05)
1,180 (110)
0.63
18.80
0.08
FE-IV
Alternative energy
0.68
0.13***
(0.04)
0.01
(0.01)
0.55
(0.60)
0.01
(0.02)
0.12
(0.33)
0.02
(0.08)
0.04
(0.09)
0.28**
(0.11)
0.11
(0.15)
0.21
(0.16)
0.10
(0.10)
0.00
(0.04)
0.00
(0.00)
608 (127)
FE-OLS
0.25***
(0.05)
0.00
(0.01)
0.62
(0.55)
0.02
(0.01)
0.56
(0.35)
0.01
(0.06)
0.06
(0.08)
0.28***
(0.09)
0.10
(0.13)
0.20
(0.14)
0.08
(0.09)
0.02
(0.04)
0.00
(0.00)
608 (127)
0.32
20.30
0.58
FE-IV
Gasoline
0.05*
(0.03)
0.00
(0.00)
0.57*
(0.30)
0.02**
(0.01)
0.34*
(0.19)
0.03
(0.04)
0.05
(0.03)
0.13*
(0.07)
0.02
(0.05)
0.01
(0.05)
0.06*
(0.03)
0.03
(0.02)
0.00
(0.00)
1,180 (110)
0.36
18.80
0.31
0.05**
(0.02)
0.00
(0.00)
0.56*
(0.32)
0.02**
(0.01)
0.33*
(0.18)
0.03
(0.04)
0.05
(0.04)
0.13*
(0.07)
0.02
(0.05)
0.01
(0.05)
0.06*
(0.03)
0.03
(0.02)
0.00
(0.00)
1,180 (110)
0.38
FE-IV
FE-OLS
Energy
0.52
0.04
(0.03)
0.00
(0.01)
0.04
(1.05)
0.01
(0.03)
0.58*
(0.32)
0.20
(0.12)
0.01
(0.04)
0.07
(0.08)
0.16
(0.12)
0.10
(0.08)
0.12
(0.10)
0.08*
(0.05)
0.00
(0.00)
1,169 (109)
FE-OLS
0.08**
(0.04)
0.00
(0.01)
0.00
(0.96)
0.01
(0.02)
0.44
(0.29)
0.20*
(0.12)
0.01
(0.04)
0.07
(0.08)
0.16
(0.11)
0.10
(0.08)
0.11
(0.09)
0.08*
(0.04)
0.00
(0.00)
1,169 (109)
0.21
18.84
0.47
FE-IV
Electricity
Note: FE estimation. All regressions use year dummies and constant (not shown). Standard errors in parentheses adjusted for within-group clustering and
heteroskedasticity. *P < 0.1; **P < 0.05; ***P < 0.01. Instruments for Kyoto: ICC membership and its spatial lag.
Polity (1 to 1)
WTO (0,1)
Openness
Gov. orientation
ln MEA stock
ln services
ln agriculture
ln manufacturing
ln population
ln GDP, squared
ln GDP
0.77
0.22***
(0.05)
0.01
(0.00)
1.30***
(0.46)
0.04***
(0.01)
0.01
(0.28)
0.04
(0.05)
0.05
(0.05)
0.22***
(0.08)
0.06
(0.06)
0.17
(0.11)
0.01
(0.06)
0.02
(0.03)
0.00
(0.00)
608 (127)
0.61
20.30
0.69
0.10**
(0.04)
0.01*
(0.01)
1.23**
(0.52)
0.04**
(0.01)
0.45
(0.28)
0.03
(0.06)
0.03
(0.06)
0.22**
(0.09)
0.07
(0.07)
0.17
(0.13)
0.03
(0.07)
0.01
(0.04)
0.00
(0.00)
608 (127)
Kyoto (0,1)
FE-IV
FE-OLS
Diesel fuel
Method
Dep. var.
Panel
Renewables
Fossil fuel
Method/sample
Long FE
(A1)
W/o EIT
(A2)
Rich only
(A3)
Long FE
(A4)
W/o EIT
(A5)
Rich only
(A6)
Kyoto (0,1)
3.43***
(1.14)
0.81***
(0.21)
4.40
(9.35)
0.29
(0.26)
7.24
(6.32)
3.06
(2.15)
2.09
(1.79)
0.44
(4.47)
1.71
(1.55)
1.11
(4.85)
2.48
(2.45)
0.40
(1.09)
0.16
(0.12)
220
110
0.32
36.49
0.32
3.44***
(1.29)
0.66***
(0.13)
6.24
(10.95)
0.06
(0.29)
4.68
(12.33)
1.43
(1.69)
2.07
(1.44)
0.09
(3.19)
0.08
(1.76)
0.42
(2.06)
0.89
(1.42)
0.32
(1.01)
0.12*
(0.06)
923
86
0.53
15.81
0.24
2.05**
(0.85)
0.07
(0.22)
5.33
(16.52)
0.24
(0.43)
0.82
(5.22)
1.31
(1.86)
0.58
(0.78)
3.37
(2.45)
1.20
(1.52)
0.71
(1.31)
0.48
(0.83)
0.32
(0.45)
0.02
(0.04)
701
65
0.72
8.39
0.05
3.62***
(1.37)
0.81***
(0.22)
8.89
(9.74)
0.41
(0.28)
2.93
(6.65)
4.25
(3.13)
3.55*
(1.90)
0.44
(4.67)
1.70
(2.04)
2.19
(5.19)
4.66*
(2.57)
0.19
(1.34)
0.13
(0.14)
220
110
0.18
36.49
0.58
2.80**
(1.29)
0.67***
(0.12)
0.74
(10.47)
0.26
(0.27)
2.30
(12.13)
0.00
(1.71)
2.49*
(1.44)
0.26
(3.30)
0.71
(1.81)
4.27*
(2.20)
1.09
(1.40)
0.44
(0.96)
0.09
(0.08)
923
86
0.50
15.81
0.12
0.58
(1.13)
0.79**
(0.40)
0.06
(18.08)
0.16
(0.45)
1.72
(6.18)
1.83
(1.96)
0.08
(0.89)
2.30
(2.75)
2.73
(2.15)
4.45***
(1.70)
0.17
(1.04)
0.15
(0.54)
0.03
(0.06)
701
65
0.58
8.39
0.02
Alternative energy
Coal
Method/sample
Long FE
(B1)
W/o EIT
(B2)
Rich only
(B3)
Long FE
(B4)
W/o EIT
(B5)
Rich only
(B6)
Kyoto (0,1)
3.06***
(0.58)
0.05
(0.06)
3.21
(4.52)
0.12
(0.13)
8.35***
(2.18)
0.28
(0.82)
0.20
(0.72)
0.16
(1.08)
1.19*
(0.69)
0.91
(2.54)
1.19
(0.87)
0.19
(0.43)
0.09
(0.09)
220
110
0.60
36.49
0.47
2.06***
(0.56)
0.00
(0.03)
1.95
(3.72)
0.10
(0.11)
0.84
(3.35)
0.78
(0.74)
0.07
(0.40)
0.87
(0.66)
0.35
(0.61)
0.20
(1.56)
0.52
(0.52)
0.33*
(0.19)
0.09
(0.06)
923
86
0.55
15.81
0.09
1.69***
(0.54)
0.12
(0.16)
2.09
(5.76)
0.00
(0.14)
1.16
(3.56)
0.63
(0.52)
0.34
(0.47)
0.16
(0.94)
0.95
(0.76)
0.26
(1.60)
0.03
(0.66)
0.00
(0.33)
0.03
(0.02)
701
65
0.30
8.39
0.21
0.96
(1.41)
0.14
(0.11)
11.34
(12.82)
0.48
(0.34)
1.89
(4.27)
6.66**
(2.59)
0.36
(2.14)
2.98
(2.13)
1.07
(2.11)
2.62
(5.91)
7.08**
(3.00)
0.92
(1.30)
0.27
(0.17)
220
110
0.62
36.49
0.74
1.71
(2.29)
0.10
(0.07)
6.34
(12.77)
0.33
(0.32)
0.66
(8.96)
3.51*
(1.84)
0.82
(1.27)
1.76
(1.31)
0.52
(1.57)
1.65
(2.81)
2.31
(1.57)
0.54
(0.83)
0.08
(0.09)
923
86
0.47
15.81
0.06
0.74
(1.30)
1.32*
(0.69)
7.10
(13.14)
0.21
(0.34)
2.29
(8.60)
2.06
(1.29)
1.51
(1.38)
0.07
(1.95)
0.88
(2.42)
1.44
(3.35)
2.28
(1.62)
0.88
(1.14)
0.37
(0.24)
701
65
0.92
8.39
0.05
Diesel fuel
Gasoline
Method/sample
Long FE
(C1)
W/O EIT
(C2)
Rich only
(C3)
Long FE
(C4)
W/o EIT
(C5)
Rich only
(C6)
Kyoto (0,1)
0.49***
(0.06)
0.01
(0.01)
0.98*
(0.54)
0.02
(0.01)
1.17***
(0.23)
0.00
(0.10)
0.06
(0.08)
0.18
(0.12)
0.11
(0.10)
0.25
(0.23)
0.13
(0.09)
0.14*
(0.07)
0.01**
(0.01)
252
126
0.04
39.06
0.47
0.12**
(0.05)
0.01
(0.00)
0.98*
(0.53)
0.03*
(0.01)
0.11
(0.39)
0.01
(0.05)
0.03
(0.05)
0.27***
(0.09)
0.08
(0.09)
0.20
(0.14)
0.04
(0.07)
0.05
(0.04)
0.00
(0.00)
498
104
0.67
18.86
0.67
0.31***
(0.08)
0.07*
(0.04)
0.67
(0.71)
0.02
(0.02)
0.07
(0.45)
0.00
(0.08)
0.14**
(0.06)
0.48***
(0.16)
0.24*
(0.13)
0.21*
(0.13)
0.06
(0.08)
0.05
(0.04)
0.01*
(0.01)
316
65
0.39
8.83
0.67
0.46***
(0.06)
0.00
(0.01)
0.55
(0.60)
0.01
(0.02)
1.34***
(0.26)
0.07
(0.10)
0.03
(0.10)
0.20
(0.14)
0.05
(0.13)
0.24
(0.21)
0.05
(0.13)
0.14*
(0.08)
0.01
(0.01)
252
126
0.47
39.06
0.40
0.17***
(0.06)
0.00
(0.01)
0.27
(0.67)
0.01
(0.02)
0.39
(0.42)
0.06
(0.09)
0.04
(0.09)
0.31***
(0.10)
0.20
(0.17)
0.26
(0.17)
0.11
(0.10)
0.03
(0.06)
0.00
(0.00)
498
104
0.34
18.86
0.56
0.32***
(0.09)
0.06*
(0.03)
0.28
(0.78)
0.01
(0.02)
0.44
(0.59)
0.02
(0.09)
0.05
(0.10)
0.51***
(0.18)
0.12
(0.12)
0.26*
(0.14)
0.11
(0.08)
0.06
(0.05)
0.01
(0.01)
316
65
0.04
8.83
0.60
Energy
Electricity
Method/sample
Long FE
(D1)
W/o EIT
(D2)
Rich only
(D3)
Long FE
(D4)
W/o EIT
(D5)
Rich only
(D6)
Kyoto (0,1)
0.06
(0.04)
0.00
(0.00)
0.73***
(0.27)
0.03***
(0.01)
0.25
(0.20)
0.03
(0.07)
0.10*
(0.05)
0.24**
(0.11)
0.03
(0.07)
0.11
(0.14)
0.13*
(0.07)
0.08**
(0.03)
0.00
(0.00)
220
110
0.67
36.49
0.02
0.07*
(0.04)
0.00
(0.00)
0.61*
(0.33)
0.02**
(0.01)
0.54**
(0.27)
0.10**
(0.04)
0.08**
(0.03)
0.19***
(0.06)
0.04
(0.07)
0.06
(0.06)
0.07*
(0.04)
0.02
(0.02)
0.00
(0.00)
923
86
0.31
15.81
0.36
0.06*
(0.03)
0.01
(0.01)
0.47
(0.38)
0.02**
(0.01)
0.38
(0.24)
0.05
(0.05)
0.01
(0.04)
0.12
(0.12)
0.04
(0.04)
0.06
(0.04)
0.00
(0.04)
0.01
(0.03)
0.01
(0.00)
701
65
0.42
8.39
0.43
0.02
(0.06)
0.00
(0.01)
0.21
(0.94)
0.02
(0.02)
0.94***
(0.25)
0.23
(0.15)
0.10
(0.08)
0.07
(0.15)
0.16
(0.14)
0.20
(0.22)
0.23
(0.20)
0.11
(0.08)
0.01
(0.01)
218
109
0.92
36.30
0.15
0.13***
(0.04)
0.00
(0.01)
1.21
(1.24)
0.02
(0.03)
0.29
(0.28)
0.37**
(0.18)
0.08
(0.05)
0.08
(0.10)
0.16
(0.19)
0.08
(0.09)
0.11
(0.11)
0.11
(0.08)
0.00
(0.00)
912
85
0.03
16.23
0.53
0.06
(0.04)
0.00
(0.01)
1.02
(0.64)
0.04**
(0.02)
0.26
(0.24)
0.00
(0.05)
0.02
(0.04)
0.11
(0.10)
0.04
(0.05)
0.06
(0.06)
0.05
(0.05)
0.03
(0.03)
0.00
(0.01)
701
65
0.74
8.39
0.65
Note: FE IV estimates. Standard errors in parentheses adjusted for within-group clustering and heteroskedasticity. All regressions use year dummies and constant (not shown). *P < 0.1; **P < 0.05; ***P
< 0.01. Instruments for Kyoto: ICC membership and its spatial lag.